Robotaxi start-up Pony.ai expects cheaper driverless cars on more roads in China in 2025

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

China’s robotaxi companies should see costs coming down and major cities opening up to their fleets next year, according to the chief executive of the latest start-up to seek more funding for its autonomous driving ambitions.

James Peng, CEO and founder of Guangzhou-based Pony.ai, said it planned to expand its robotaxi fleet from about 250 to at least 1,000 vehicles in 2025, with lower production costs and larger service areas in the so-called first-tier cities of Beijing, Guangzhou, Shanghai and Shenzhen. 

“Technological advances will allow us to reduce costs of production by several times,” Peng said in a Financial Times interview, adding that the first-tier cities would “soon” open up significantly, if not entirely, to driverless taxis.

The founder said this could mean Pony.ai reporting a positive margin for its robotaxi business as soon as next year.

However, investors are unconvinced about the sector’s prospects, with the company and its peers all lossmaking. At the end of November, Pony.ai joined a roster of self-driving start-ups going public this year. It raised $452mn in a Nasdaq offering and through private placements, but its shares fell nearly 8 per cent upon their New York debut.

The lacklustre float highlights market scepticism about whether the industry can become commercially viable amid fierce competition, an uncertain policy outlook, heavy spending on research and development and sparse revenues in the short term. In the US, General Motors abandoned the development of its Cruise robotaxi business this month.

The Nasdaq IPO valued Pony.ai at $5.25bn, nearly 40 per cent lower than the $8.5bn it was worth two years ago. Local peers WeRide and Horizon Robotics saw their valuations fall from previous funding rounds by 22 per cent and 23 per cent, respectively, when they sold shares in October.

Pony.ai’s driverless taxi services in China’s first-tier cities have been slow to take off due to its small fleet and limited areas of service, covering only a handful of districts. “We don’t have enough users because we don’t have enough robotaxis on the road,” said Peng. 

The company derives more than two-thirds of its revenues — or $27.4mn of $39.5mn in the first nine months of this year — from providing driverless truck services, but Peng said that could change soon, with the start-up partnering with two Chinese state-owned carmakers to “mass produce thousands of” robotaxis a year.

The initiative will allow Pony.ai to lower production costs to less than Rmb300,000 ($41,000) a car through economies of scale, said Peng. It currently spends more than Rmb500,000 making a driverless taxi, according to people close to the company.

Meanwhile, Peng expects Beijing to allow driverless taxis to roam all suburban districts within the next two years and Guangzhou, a commercial hub, to open up the bulk of the city over the same period.

His chief technology officer, Lou Tiancheng, said Pony.ai did not plan to target the lower end of the market, where margins are lower and demand is stronger, but was aiming to attract customers “willing to pay a premium price for a better experience”.

“Your first two orders [with us] may be cheap,” he said. “It will not be the case for your future orders.”

Lou added that the company could “make a lot of money” when its robotaxi fleet grew to 10,000, an “achievable” target that only accounted for “a very small percentage” of China’s taxi fleets of all varieties.

Analysts, however, are not convinced. Tu Le, founder of Sino Auto Insights, a Detroit-based consultancy, said he was “impressed” by Pony.ai’s robotaxi after riding it this year, but he was not sure how it could stand out in a market with “immense competition” in which it lagged behind more established peers in everything from capital to fleet size and cost control.

“We still don’t know what is the secret sauce that makes Pony better than everybody else,” said Le.

Regulatory uncertainty could create another hurdle for Pony.ai’s expansion. While major cities are embracing the technology, smaller ones are more cautious. Lou said his company would not consider entering certain cities even after they unveiled policies to encourage self-driving.

“We would love to enter a market with very strong government support,” he said.

Chief executive Peng insisted the company was working in a field with “not many” competitors. He said he was not concerned about Tesla’s plan to expand into robotaxis, as the electric vehicle giant had not “invested long enough” in the sector.

He added that Tesla might face “challenges” in winning regulatory approval to launch its signature semi-autonomous driving service in China.

“It would be a good thing if we let [the driving service] in,” he said. “It is better to have competition.”

Additional reporting by Gloria Li in Hong Kong